A method of estimating the yield of a missing plot in field experimental work

In cases of field experiments when one plot is missing, a method has been developed for furnishing an estimate of the yield of the missing plot, based on all the other values. The calculation is given for (a) a Randomised Block experiment, and (b) a Latin Square arrangement. In both cases the actual arithmetic is very simple.The steps in the procedure are:(1) Determine the desired value by an application of the equation (A) or (B) according as the experiment was arranged in Randomised Blocks or in a Latin Square.(2) Proceed as usual with the analysis of variance, using the estimated figure for the missing yield, and remembering to deduct one from the number of degrees of freedom ascribable to error.